Ridge Regression for Addressing of the Multicollinearity Problem with Application in Cost of Production

Ali Sadig Mohommed BAGER; Bahr KADHIM MOHAMMED; Meshal HARBI ODAH

doi:10.18662/lumproc.nashs2017.4

Ali Sadig Mohommed BAGER The Bucharest University of Economic Studies, Department of Statistics and Econometrics, Iraq, Muthanna University
Bahr KADHIM MOHAMMED The Bucharest University of Economic Studies, Department of Statistics and Econometrics, Iraq, University of AL-Qadisiyah
Meshal HARBI ODAH The Bucharest University of Economic Studies, Department of Statistics and Econometrics, Iraq, Muthanna University

DOI: https://doi.org/10.18662/lumproc.nashs2017.4

Keywords: Ridge regression, Mullticollinearity problem, Production of cement

Abstract

The regression analysis is statistical method of extensive use, which illustrates the relationship between the explanatory variables and the dependent variable in the form of a model useful in the interpretation of scientific phenomenon, bringing also benefits to society. In this paper we study the most important factors affecting the cost production of cement (Muthanna Factory) by using the ridge regression. The factors are described as follows: we consider the cost of production amount as response variable and factors that affect or may affect the explanatory variables are labor, Price per ton, Electric power, Quantity consumed. They all suffer from high correlation, indicating a problem of multicollinearity .The data analysis is included in the study of the ridge regression as the best approach in case of a multicollinearity problem in the context of financial and economic data being associated with each other often. We used R packages (MASS).

References

[1] Al-Jubouri & Habib C. "Multiple Regression and Analysis of Differentiation", (translated), Iraq: Higher Education Printing Press. 1990
[2] Dorugade, A.V. New ridge parameters for ridge regression. Journal of the Association of Arab Universities for Basic and Applied Sciences. 2014 (15). pp. 94-99
[3] Anwar F. & Lee Ceng Y. "Performance of Ridge Regression Estimator Method on Small Sample size By Varying correlation coefficients: A simulation study", Journal of Mathematics and statistics 10 (1). 2014. pp. 25 – 29
[4] García C.B., García, J., López Martín, M.M., & Salmerón, R.. Collinearity: Revisiting the variance inflation factor in ridge regression. Journal of Applied Statistics, 42(3), 2015. pp. 648-661
[5] Hoer, A.E, & Kennard R.W. Ridge Regression. Advances, Algorithms and Applications 1981: American Sciences Press. 1980
[6] Hoerl A.E., Kannard R.W., & Baldwin K.F. Ridge regression: some esimulations. Communications in Statistics-Theory and Methods. 4 (2). 1975. pp. 105-123
[7] Hoerl A.E., & R W. Kennard. "Ridge regression: Iterative estimation of the biasing parameter," Commun. Statist.A5. 1976. pp. 77-88
[8] Kibria, B.G. Performance of some new ridge regression estimators. Communications in Statistics-Simulation and Computation. 2003. 32 (2). pp. 419-435
[9] Kraha A., Turner H., Nimon K., Zientek L. R., & Henson R. K. Tools to support interpreting multiple regression in the face of multicollinearity. Frontiers in psychology. 3. 2012
[10] Kazem, Armory H. & Shalibah, M.B. "Advanced Economic Measurement Theory and Practice", Baghdad: Dunia al-Amal Library. 2002
[11] McDonald G.C. Ridge regression. Wiley Interdisciplinary Reviews: Computational Statistics, 1(1). pp. 93-100. 2009